7 ms·

Remotely controlled humanoid robots are pretty cool though, aren’t they? Like, potentially hugely valuable?

dom96 2y ago
Yeah, they were cool in the 2000s too when ASIMO was shown off. What ever happened to that? Honda is now valued higher than Tesla, right? Oh wait... no they are not.
- chrisco255 2y ago
  ASIMO never made it off stage, and Honda never prioritized shipping it. We definitely never saw ASIMO milling around a crowded event and serving drinks and handing out gift bags. It was never remote controllable by unskilled operators. It doesn't take a whole lot of imagination to come up with scenarios where even a remote controlled bot would be incredibly useful to society. If Honda never productized it or developed Asimo into something practical, then we can only speculate the reasons why. But that's on Honda for failing to follow through, ultimately.
mossTechnician 2y ago
Generally, humanoid robots are not preferable to other, purpose-built robots that (for example) you'll see on factory floors. Boston Dynamics already have adaptable, quadrupedal robots that can be remotely deployed too. The tasks Tesla suggested included babysitting, and now that we know that the robots are built for remote control first, I would feel pretty uncomfortable with letting my child alone in the same house with a robot like that. Vacuum cleaners are bad enough. https://www.vice.com/en/article/ecovacs-robot-vacuums-hacked-slurs/ https://www.vice.com/en/article/ecovacs-robot-vacuums-hacked...
- schiffern 2y ago
  >remote control first If you mean 'first' in a literal chronological sense, sure. It's a logical R&D stepping stone. If you mean it's the typical operating mode for the final product, I doubt we can conclude that just from an early public tech demo.
  - mossTechnician 2y ago
    Touche; I don't have a reason to think any finished product would be remote control first, but I would be shocked if the resulting robot wasn't bristling with cameras and network antennas. If hackers can yell slurs and record people through a robot vacuum, how much more could they do through a robot butler?
- jncfhnb 2y ago
  That is because humanoid robots are general purpose forms, rather than purpose built forms. Factory robotics is mostly just conveyor belts and pusher whatevers. Not even something resembling a droid of any kind.
trzy 2y ago
Would you pay $20k + $3/hr for a teleoperated housekeeer that works slowly instead of a $20/hr human? It just doesn’t make any financial sense. Automation is the only way but even then, at say $200-$300/mo, the break even is ~2 years on a $5000 robot, and that’s ignoring maintenance costs. There are also potential ethical issues around teleoperation as a form of labor arbitrage. Should employers be able to outsource low-skill physical labor to actual humans abroad to circumvent wage and immigration laws? Is the benefit worth the economic dislocation to society? If you say yes, ask yourself why even bother reshoring manufacturing at higher cost and with trade barriers. Teleoperation is argued to be an interim solution that trains autonomous systems the way driving trains FSD but as of right now, given what we know about just how much data is needed for current imitation learning approaches to work for fine manipulation tasks, it doesn’t seem scaleable. Maybe there will be a modeling breakthrough. My bet is on cheaper non-humanoid form factors and data collection that doesn’t require a robot being operated for every data collector in the field.
- fakedang 2y ago
  Mining. War. Sewer cleaning. Tons of jobs where it would help not being actively in the call of duty, and where precision isn't as much of a priority.
  - trzy 2y ago
    Teleop is used already in these applications. But a human form factor is not required.
    - jncfhnb 2y ago
      A human form factor is not required or optimized for any specific task, but it does generalize which is potentially very useful.
      - mckn1ght 2y ago
        Maybe. But if it is a worse design for tasks A, B and C and harder to build, why not go with the simpler and more specialized design for each separate task? Is the economy of scale really that much better with the general design? I’m not convinced.
underwater 2y ago
The creators of Mechanical Turk no doubt made money from it in 1770. https://en.m.wikipedia.org/wiki/Mechanical_Turk https://en.m.wikipedia.org/wiki/Mechanical_Turk
- swores 2y ago
  Although that was neither remotely controlled, nor humanoid, nor a robot.

Yeah, they were cool in the 2000s too when ASIMO was shown off. What ever happened to that? Honda is now valued higher than Tesla, right? Oh wait... no they are not.

chrisco255 2y ago
ASIMO never made it off stage, and Honda never prioritized shipping it. We definitely never saw ASIMO milling around a crowded event and serving drinks and handing out gift bags. It was never remote controllable by unskilled operators. It doesn't take a whole lot of imagination to come up with scenarios where even a remote controlled bot would be incredibly useful to society. If Honda never productized it or developed Asimo into something practical, then we can only speculate the reasons why. But that's on Honda for failing to follow through, ultimately.

mossTechnician 2y ago

Generally, humanoid robots are not preferable to other, purpose-built robots that (for example) you'll see on factory floors. Boston Dynamics already have adaptable, quadrupedal robots that can be remotely deployed too. The tasks Tesla suggested included babysitting, and now that we know that the robots are built for remote control first, I would feel pretty uncomfortable with letting my child alone in the same house with a robot like that. Vacuum cleaners are bad enough. https://www.vice.com/en/article/ecovacs-robot-vacuums-hacked-slurs/ https://www.vice.com/en/article/ecovacs-robot-vacuums-hacked...

schiffern 2y ago
>remote control first If you mean 'first' in a literal chronological sense, sure. It's a logical R&D stepping stone. If you mean it's the typical operating mode for the final product, I doubt we can conclude that just from an early public tech demo.
- mossTechnician 2y ago
  Touche; I don't have a reason to think any finished product would be remote control first, but I would be shocked if the resulting robot wasn't bristling with cameras and network antennas. If hackers can yell slurs and record people through a robot vacuum, how much more could they do through a robot butler?
jncfhnb 2y ago
That is because humanoid robots are general purpose forms, rather than purpose built forms. Factory robotics is mostly just conveyor belts and pusher whatevers. Not even something resembling a droid of any kind.

trzy 2y ago

Would you pay $20k + $3/hr for a teleoperated housekeeer that works slowly instead of a $20/hr human? It just doesn’t make any financial sense. Automation is the only way but even then, at say $200-$300/mo, the break even is ~2 years on a $5000 robot, and that’s ignoring maintenance costs. There are also potential ethical issues around teleoperation as a form of labor arbitrage. Should employers be able to outsource low-skill physical labor to actual humans abroad to circumvent wage and immigration laws? Is the benefit worth the economic dislocation to society? If you say yes, ask yourself why even bother reshoring manufacturing at higher cost and with trade barriers. Teleoperation is argued to be an interim solution that trains autonomous systems the way driving trains FSD but as of right now, given what we know about just how much data is needed for current imitation learning approaches to work for fine manipulation tasks, it doesn’t seem scaleable. Maybe there will be a modeling breakthrough. My bet is on cheaper non-humanoid form factors and data collection that doesn’t require a robot being operated for every data collector in the field.

fakedang 2y ago
Mining. War. Sewer cleaning. Tons of jobs where it would help not being actively in the call of duty, and where precision isn't as much of a priority.
- trzy 2y ago
  Teleop is used already in these applications. But a human form factor is not required.
  - jncfhnb 2y ago
    A human form factor is not required or optimized for any specific task, but it does generalize which is potentially very useful.
    - mckn1ght 2y ago
      Maybe. But if it is a worse design for tasks A, B and C and harder to build, why not go with the simpler and more specialized design for each separate task? Is the economy of scale really that much better with the general design? I’m not convinced.

underwater 2y ago

The creators of Mechanical Turk no doubt made money from it in 1770. https://en.m.wikipedia.org/wiki/Mechanical_Turk https://en.m.wikipedia.org/wiki/Mechanical_Turk

swores 2y ago
Although that was neither remotely controlled, nor humanoid, nor a robot.